Electromagnetic field radiated by a point emitter on a graphene sheet
Electromagnetic fields radiated from electrostatic ...
Transcript of Electromagnetic fields radiated from electrostatic ...
A111D2 7fi3flDa
NAFL INST OF STANDARDS & TECH R.I.C.
A1 11 02783808/Electromaanetlc fields radiated from elQC100 .U5753 N0.1314 1988 V198 C.I NBS-P
NBS
PUBLICATIONS
••'•tiiu 0** NBS TECHNICAL NOTE 1314
U.S. DEPARTMENT OF COMMERCE / National Bureau of Standards
Electromagnetic Fields
Radiated From Electrostatic Discharges
Theory and Experiment
Perry F. Wilson
Arthur R. Ondrejka
Mark T. MaJohn M. Ladbury
-QC
100
.U5753
#1314
1988
C.2
Tm he National Bureau of Standards' was established by an act of Congress on March 3, 1901. The Bureau's overall
^j^^ goal is to strengthen and advance the nation's science and technology and facilitate their effective application for
public benefit. To this end, the Bureau conducts research to assure international competitiveness and leadership of U.S.industry, science arid technology. NBS work involves development and transfer of measurements, standards and related
science and technology, in support of continually improving U.S. productivity, product quality and reliability, innovation
and underlying science and engineering. The Bureau's technical work is performed by the National MeasurementLaboratory, the National Engineering Laboratory, the Institute for Computer Sciences and Technology, and the Institute
for Materials Science and Engineering.
The National Measurement Laboratory
Provides the national system of physical and chemical measurement;
coordinates the system with measurement systems of other nations and
furnishes essential services leading to accurate and uniform physical andchemical measurement throughout the Nation's scientific community,industry, and commerce; provides advisory and research services to other
Government agencies; conducts physical and chemical research; develops,
produces, and distributes Standard Reference Materials; provides
calibration services; and manages the National Standard Reference Data
System. The Laboratory consists of the following centers:
• Basic Standards^• Radiation Research• Chemical Physics• Analytical Chemistry
The National Engineering Laboratory
Provides technology and technical services to the public and private sectors
to address national needs and to solve national problems; conducts research
in engineering and applied science in support of these efforts; builds andmaintains competence in the necessary disciplines required to carry out this
research and technical service; develops engineering data and measurementcapabilities; provides engineering measurement traceability services;
develops test methods and proposes engineering standards and code
changes; develops and proposes new engineering practices; and develops
and improves mechanisms to transfer results of its research to the ultimate
user. Tlie Laboratory consists of the following centers:
Applied Mathematics
Electronics and Electrical
Engineering^
Manufacturing Engineering
Building TechnologyFire Research
Chemical Engineering^
The Institute for Computer Sciences and Technology
Conducts research and provides scientific and technical services to aid
Federal agencies in the selection, acquisition, application, and use of
computer technology to improve effectiveness and economy in Governmentoperations in accordance with Public Law 89-306 (40 U.S.C. 759),
relevant Executive Orders, and other directives; carries out this mission bymanaging the Federal Information Processing Standards Program,developing Federal ADP standards guidelines, and managing Federal
participation in ADP voluntary standardization activities; provides scientific
and technological advisory services and assistance to Federal agencies; andprovides the technical foundation for computer-related poUcies of the
Federal Government. The Institute consists of the following divisions:
Information Systems Engineering
Systems and Software
TechnologyComputer Security
Systems and NetworkArchitecture
Advanced Computer Systems
The Institute for Materials Science and Engineering
Conducts research and provides measurements, data, standards, reference
materials, quantitative understanding and other technical informationfundamental to the processing, structure, properties and performance ofmaterials; addresses the scientific basis for new advanced materials
technologies; plans research around cross-cutting scientific themes such as
nondestructive evaluation and phase diagram development; oversees
Bureau-wide technical programs in nuclear reactor radiation research andnondestructive evaluation; and broadly disseminates generic technical
information resulting from its programs. The Institute consists of the
following Divisions:
• Ceramics• Fracture and Deformation^• Polymers• Metallurgy• Reactor Radiation
'Hcadquancrs and Laboratories at Gaithersburg, MD, unless othei^vise noted; mailing address
Gaithcrsburg, MD 20899.
'Some divisions within the center are located at Boulder, CO 80303.'Located at Boulder, CO, with some elements at Gaithersburg, MD
Research Information Center
(National tiureau of Standards
Gaithersburg, Maryland 20899
Qjlloo
Electromagnetic Fields Radiated From ^^7^3
Electrostatic Discharges—Theory mf^and Experiment
Perry F. Wilson*
Arthur R. Ondrejka
Mark T. MaJohn M. Ladbury
Electromagnetic Fields Division
Center for Electronics and Electrical Engineering
National Engineering Laboratory
National Bureau of Standards
Boulder, Colorado 80303-3328
*Presently with BBC, Baden-Dattwil, Switzerland
U.S. DEPARTMENT OF COMMERCE, C. William Verity, Secretary
NATIONAL BUREAU OF STANDARDS, Ernest Ambler, Director
Issued February 1988
National Bureau of Standards Technical Note 1314
Natl. Bur. Stand. (U.S.), Tech Note 1314, 72 pages (Feb. 1988)
CODEN:NBTNAE
U.S. GOVERNMENT PRINTING OFFICEWASHINGTON: 1988
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402
CONTENTS
List of FiguresAbstract
1
.
INTRODUCTION
2
.
PRE-DISCHARGE EFFECTS AND DISCHARGE CURRENT
3
.
DIPOLE MODEL OF THE ESD -RADIATED FIELDS
3 .
1
Electric Field
3 .
2
Magnetic Field
4. MEASURED ESD CURRENT AND ELECTRIC -FIELD WAVEFORMS
4 .
1
Instrximentation ,
4 .
2
Experimental Results ,
5
.
CONCLUDING REMARKS AND DISCUSSIONS
6
.
ACKNOWLEDGMENTS
7 REFERENCES
APPENDIX A: Time-Dependent Fields Due to a Dipole Source
APPENDIX B: Spectrum of ESD Dipole Radiation Model
APPENDIX C: Time-Domain Antennas for Measuring ImpilsiveElectromagnetic Fields
Page
V1
2
3
5
6
8
8
8
10
14
15
15
17
24
26
iii
List of Figures
Figure Number Page
1. Diagrams showing the development of a spark channel,based on cloud chamber observations: (a) electronavalanche, (b) channel growth toward anode, (c) channelgrowth toward both anode and cathode, and (d) conductivebridge [7] 27
2. Conventional single RC circuit model for humanelectrostatic discharge [9] 28
3. The dual RLC circuit model for human electrostaticdischarge incorporating separate, parallel paths forbody and hand discharges [11] 28
4. Computer simulation of an ESD current based on the
dual RLC circuit (fig. 3). The circuit parameters usedare C = 7.5 pF, R^ = 200 fi, L^ = 0.1 /iH, C = 100 pF,
R_ = 500 Q, L„ = 1.7 /xH, and 5-kV initial charge [11] 29
5. Computer solutions for a dual RLC-circuit model usingdifferent single-pole simulated oscilloscope bandwidths
.
The initial applied voltage is 5 kV [15] 30
6. The raidation spectrum at 2.0 m from an ESD sourcefor various applied voltages [1] 31
7. A dipole above a perfect ground plane and its image 32
8
.
Measurement system schematics 33
(a) For measuring current waveform.(b) For measuring electric field.
9. Coaxial ball target measurement configuration(Shield material was used to prevent radiation ofthe ESD simulator from interfering with thereceiving system) 34
10. Coaxial ball target current and vertical electricfield for a 1-kV discharge 35
11. Coaxial ball target current and vertical electricfield for a 2-kV discharge 36
12. Coaxial ball target current and vertical electricfield for a 4-kV discharge 37
13. Coaxial ball target current and vertical electricfield for a 6-kV discharge 38
V
14. Coaxial ball target current and vertical electricfield for a 8-kV discharge 39
15. Coaxial ball target current and vertical electricfield for a 10-kV discharge 40
16. Coaxial ball target current and vertical electricfield for a 12 -kV discharge 41
17. Predicted vertical electric field for the 4-kVdischarge of fig. 12 as a function of distance to
the observation point 42
18. Predicted magnetic field for the 4-kV discharge case: 43(a) with the receiving antenna at 1.5 m from the
spark, and (b) as a function of the distance to theobservation point.
19. Spectral content for the coaxial ball target 4-kVdischarge of fig . 12 '. 44
20. Vertical electric field for a 1-kV spark to a
ground plane 45
21. Vertical electric field for a 2-kV spark to a
ground plane 46
22. Vertical electric field for a 4-kV spark to a
ground plane 47
23. Vertical electric field for a 6-kV spark to a
ground plane 48
24. Measurement configuration for indirect radiationfrom a vertical square metal plate 49
25. Vertical electric field radiated by a verticalsquare metal plate excited by a 5-kV spark 50
26. Vertical electric field radiated by a verticalsquare metal plate excited by a 7-kV spark 51
27. Measurement configuration for radiation by a metalchair above a ground plane 52
28. Vertical electric field radiated by a metal chairabove a ground plane excited by a 3-kV spark 53
29. Vertical electric field radiated by a metal chairabove a ground plane excited by a 6-kV spark 54
30. Measurement configuration for radiation by a metaltrash can on a ground plane 55
VI
31. Vertical electric field radiated by a metal trash canexcited by a 4-kV spark 56
32. The UHF antenna (NBS 200-1) for measuring the
broadband electric field: dimensions in the figureare in centimeters 57
33. Construction details of the pulse transformer in theNBS 200-1 antenna (all dimensions in centimeters) 58
34. Comparison of the frequency response of the NBS 200-1
antenna with and without the Kydex jacket 59
35. Normalized amplitude of the transmitting pulse 60
36. Response of the NBS 200-1 antenna to the transmittingpulse shown in fig. 35 60
37. The microwave horn antenna, NBS 50-2 61
38. The broadband balun used for the antenna in fig. 37 62
VI
1
ELECTROMAGNETIC FIELDS RADIATED FROM ELECTROSTATIC DISCHARGESTHEORY AND EXPERIMENT
Perry F. Wilson*, Arthur R. Ondrejka, Mark T. Ma, and John M. LadburyElectromagnetic Fields DivisionNational Bureau of Standards
Boulder, Colorado 80303
The fields radiated by electrostatic discharges (ESD) arestudied both theoretically and experimentally. The ESD sparkis modeled theoretically as an electrically short, timedependent, linear dipole situated above an infinite groundplane. Experimentally, sparks of varying voltages aregenerated by a commerically available simulator and used to
excite a number of targets including (1) the extended innerconductor of a coaxial cable mounted in a ground plane, (2)
direct discharges to a ground plane, (3) indirect radiationfrom a large metal plate, (4) a metal chair over a groundplane, and (5) a metal trash can. Results show thatrelatively low-voltage sparks (2-4 kV) excite the strongestradiated fields. This suggests that the spark fields can posea significant interference threat to electronic equipment intothe gigahertz range.
Key words: electrostatic discharge; pulsed current; radiatedelectric field; radiated magnetic field; spectrum; time-domaintechnique
.
*Presently with BBC, Baden- Dattwil , Switzerland.
1. INTRODUCTION
Electrostatic discharge, or ESD, is a common phenomenon with the
potential to seriously upset electronic equipment. Currently, electronicsmanufacturers and users spend large sums for ESD prevention. As a result,
ESD testing is now a widespread practice. Considerable effort has gone into
the study of the ESD currents to develop current simulators for simplifyingand standardizing ESD testing. However, the ESD-radiated fields havereceived significantly less attention, and only a few measurements haveappeared in the literature [1]. This is primarily because fast field pulsesare difficult to measure with available antennas. The recent development bythe National Bureau of Standards (NBS) of broadband, linear-phase, time-
domain antennas gives us the necessary measurement capability. This reportexamines theoretically and experimentally ESD- related fields and seeks to
better quantify their interference threat.
As the name implies, ESD consists of two different phases.Electrostatic refers to the relatively slow buildup of a local voltagedifferential and the associated static electric field. Discharge denotesthe rapid transfer of the resultant accumulated charge due to the breakdownof an intervening insulator, typically air. The basic ESD phases are [2]:
(1) the corona- generated rf fields, (2) the pre-discharge effects, (3) the
discharge current, (4) the discharge electric field, and (5) the dischargemagnetic field. Each can affect sensitive electronic equipment. Thecorona, pre-discharge effects, and channel development are briefly discussedin Section 2. However, the main threat comes from the currents and fieldsthat result from actual discharge.
The ESD fields problem will be considered analytically in Section 3
based on a simple dipole model. The spark is modeled as an electricallyshort, time - dependent , linear source (or dipole) situated above an infiniteground plane. In practice, discharges are usually to some metal object suchas an equipment case or metal cart. If we allow the dipole to be situatedclose to the ground plane, then locally this should approximately reproduceESD field behavior. This model allows us to examine the electric andmagnetic near- field components, either on the ground plane surface whichaffects direct penetration into equipment shielding, or radiated away fromthe ground plane as in the case of interference to nearby electronics. Themodel is suitable for experimental confirmation and parametric studies.
Section 4 discusses the measurement system and the experimental data.
The measurement of unknown pulsed currents and fields requires that the
system be carefully calibrated with known pulses. The system response may
then be deconvolved out of the measured data to yield the true measuredpulses. The details of this procedure as well as the system limitations are
presented in Section 4.1. The test results are covered in Section 4.2. The
basic procedure is to excite a metal object with a spark and measure the
radiated electric field. The objects excited are (1) the extended inner
conductor of a coaxial cable mounted in a ground plane, (2) direct
discharges to a ground plane, (3) indirect radiation from a large, vectical,
metal plate, (4) a metal chair over a ground plane, and (5) a metal trash
can. We find that it is the low-voltage (2 - 4 kV) , fast rise-time (less
than 1 ns) sparks which produce the highest radiated electric field. This
result is also expected from the theory, which predicts dependence of the
far- field radiation on the derivative of the current waveform. These fields
often exceed 100 V/m at a distance of 1.5 m from the discharge point andclearly pose an interference threat to nearby electronics. The coaxialtarget also allows us to measure the current waveform which is then used in
the theoretical model described in Section 3 to predict the radiated fields(both electric and magnetic). The agreement between theory and experiment,for the electric field only, is quite good considering the idealized natureof the model. The magnetic fields are not measured due to the lack of a
suitable broadband H- field antenna.
We conclude with summary of the results and suggestions for furtherresearch.
2. PRE-DISCHARGE EFFECTS AND DISCHARGE CURRENT
Objects accumulate charge either triboelectrically or inductively.Triboelectric charging is a mechanical process whereby relative surfacemotion transfers charge. An example is a person walking on a carpet.Charge transfer depends on the amount of contact, surface smoothness, airhumidity, contact pressure, the triboelectric properties of the rubbingmaterials, and the rate of relative motion [3]. Precautionary measures to
prevent excess charging are an important step in ESD prevention. Theseinclude anti-static carpet, anti-static mats, and control of humidity.Despite these measures, charged persons and carts are probably unavoidable.
The voltage to which a person or cart can be charged is highlydependent on their capacitance. Human capacitance can be modeled usingcylinders and spheres. Based on this type of calculation, a typical figurefor the lumped capacitance of an adult slightly above ground is 100 pF [4] .
Common voltage figures quoted are 8-10 kV [5] for a charged person, withsome authors claiming voltages as great as 30 kV. However, corona bleed- offlikely keeps the upper bound more on the order of 25 kV. Persons typicallydo not feel sparks due to voltages less than 3 kV. Thus, many potentiallyupsetting ESD events go undetected.
Induction charging is the result of exposing an ungrounded object to anelectrostatic field [3]. This often happens during the packaging andtransportation of electronics. Handling, conveyor belts and other frictionsources can cause a substantial charge buildup on a box or plastic package.The resulting local electrostatic field can affect solid-state logic throughbit reversal, charge drainage, and dielectric burn through [6].
Under normal conditions, a gas such as air is a very effectiveinsulator. Electrons are tightly bound and current conduction is difficult.Atoms are free to move about but are constrained by collisions. Collisionrates and speeds are determined by the gas type, pressure, and temperature.Free electrons and positive ions may be present, but their numbers are
usually very low and their motion will be random. Thus, on average, there
is no transfer of charge. Applying an electrostatic field introduces a biasto the movement of charged particles. Electrons will now attempt to flowtoward the anode while positive ions will flow toward the cathode.Electrons are lighter than the positive ions (at a ratio of 1:1800) and havesmaller collision cross sections. Therefore, electrons contribute the bulkof the current. High- field levels can lead to the development of a very low
impedance channel within the gas resulting in a rapid charge transfer. Theprocess, referred to as the electrical breakdown of gas, is complicated andits study has been an important research area for over a century.
ESD depends on the creation of free electrons and the formation of a
conductive channel, as sequenced in figure 1, which is based on cloudchamber observations [7]. In (a) electrons released from the cathodecollide to form additional electrons, in effect creating an avalanche in the
direction of the anode. As the electron avalanche progresses it leavesbehind positive ions, the space charge or corona cloud, which sets up a
field perpendicular to the applied field. This tends to keep the avalanchein a well-defined channel (b) . The conductive channel will continue to growtoward the anode due to the applied field and also toward the cathode due to
the large trailing space charge (c) . When the channel reaches bothterminals (d) a conductive bridge is formed and a large current (spark) is
possible. The discharge rapidly drains the available charge and when the
current drops below some critical value, the conductive channel fails.
The pre-discharge current i may be estimated by [ 8
]
i = VCv (1)P P
where V is the applied voltage, C is the leader capacitance (on the order of5 pF/cm) , and v is the velocity at which the leader propagates. The
propagation velocity depends primarily on the applied voltage. For example,a value of 0.05 cm/ns at 35 kV is quoted by Rhoades [8]. However, leadervelocities for arbitrary voltages are not well known. Using these numbersfor a 1.0 -cm gap, we see that the pre-discharge current would be 8.75 A, andlast for about 20 ns . Typical breakdown voltages are significantly less.Thus, a level of 8.75 A for the pre-discharge current may be considered anupper bound. As will be seen in Section 4, a current building up over a 20
ns period is relatively slow compared to the actual discharge current and is
not expected to radiate significant fields. Higher voltages mean largerspark gaps and consequently long time to form the conductive channel. Theresult is a longer current rise time for higher applied voltages. Becausemany ESD hazards are related to the rate of change of the discharge current,equipment will sometimes pass high voltage tests and fail at lower dischargevoltages. This counter- intuitive effect has been noted frequently in the
literature, but is not yet universally recognized.
The pre-discharge fields and currents may pose a threat to sensitiveelectronics. However, because the actual discharge currents and fieldsassociated with them are significantly higher, the pre-discharge effects are
largely ignored when analyzing ESD hazards.
The discharge current depends on the RLC characteristics of the chargesource, the charging voltage, the rate of approach leading to the discharge,the electrode geometry, the conductive channel characteristics, and the RLCcharacteristics of the discharge victim. Given the wide ranges these
variables can cover, it is difficult to construct a simple, general model of
an ESD event. However, efforts are underway by the IEEE, the lEC, and
others to develop standard current shapes, targets, and test configurationsin order to make ESD testing more repeatable and meaningful.
For long-term ESD testing it is far too difficult and variable for
persons to repeatedly charge and discharge themselves. Thus, most ESD
testing is done with simulators. This raises the problem of how to designan RLC circuit which realistically simulates human characteristics. Initialproposals, such as the current version of lEC Publication 801-2 [9], suggesta simple RC circuit as shown in figure 2. Values proposed by variousorganizations range from 60 to 300 pF for C and 10 to 10000 Q for R [10].
However, this model ignores the inductance inherent to a human discharge andis too simple to yield realistic results.
Richman [10-11] has proposed a dual RLC circuit, as shown in figure 3.
This model recognizes that the hand- forearm (Iv,,Ru,C ) and body (L , R , C )
have very different discharge characteristics and therefore should not belumped together. A further refinement is that proposed by Boxleitner [5].
He adds a finger capacitance which modifies the initial spike, dischargepath impedances, and lumped impedances for the discharge victim, in additionto the arm and body impedances of the dual LRC circuit.
A computer simulation of a discharge based on the dual RLC circuitmodel in figure 3 is shown in figure 4. The hand- forearm contributiondischarges quickly and contributes the fast rise -time initial spike. Thebody circuit generates the longer tail as the bulk of the stored charge is
dissipated. In this particular example, the rise time is approximately 0.7
ns . If lower inductances (than 0.1 /xH) were used, still faster rise timeswould result. A much faster rise time of the discharge current has alsobeen observed by actual measurements when a simulator is approached to a
testing object with a higher speed [12-14].
The measurement of current shapes is very dependent on the bandwidth ofthe oscilloscope used, as well as the characteristics of the victim [15].Initial measurements often used fairly slow oscilloscopes (less than 100 MHzbandwidth) and the very fast rise times possible for low voltage sparks werenot immediately recognized. Thus, the contribution due to the hand- forearmwas overlooked and the single RLC circuits resulted. At present, it is
recognized that oscilloscopes with bandwidths of at least 1 GHz arenecessary to characterize current waveforms. A computer simulation byRichman [15] illustrating how oscilloscope bandwidth affects the measuredcurrent waveform is given in figure 5. Our measurement system has aneffective bandwidth on the order of 1 GHz.
3. DIPOLE MODEL OF THE ESD-RADIATED FIELDS
The fields due to ESD currents can penetrate equipment directly or
excite apertures, seams, vents, input-output cables, and the like, andcouple to susceptible internal circuitry. Thus, the designer seeks to
minimize this coupling at frequencies where problems are expected to occur.
Reciprocity dictates that inefficient receivers will also be poor radiators.Therefore, equipment which can pass stringent ESD tests are usually withinFCC- type specifications for radiated EMI limits as well.
3.1 Electric Field
Electric fields tend to excite high- impedance antennas and voltage-sensitive circuitry. Unwanted coupling can therefore be reduced by keepingthe impedance of any potential receiving antenna low. This may enhancemagnetic field pickup, so there are typically trade-offs. The basic rule is
to keep lines short. Boxleitner [5] presents a detailed discussion on howto lay out circuit boards to minimize line-antenna problems.
Honda and Ogura have performed field measurements for various idealizedESD events [1]. They charged a pair of colinear conductive cylinders onesuspended above the other and then drop the upper cylinder to bring themtogether and create a spark. An example of their data is shown in figure 6.
They measure a fairly broadband spectrum which widens with increasingapplied voltage up to about 10 kV and then begins to decrease as coronaincreasingly dissipates the energy before the main discharge. Their dataindicate that most of the ESD energy is at frequencies below 2 GHz.However, ESD spectral distribution will depend on the current waveform,conductor geometry, measurement position etc., in addition to the chargingvoltage. Thus, although the data in figure 6 are likely representative,much work remains toward well characterizing ESD in the frequency domain.
We will model an ESD discharge by a dipole of length di above a perfectground, as shown in figure 7. We impose a cylindrical coordinate system(p,(f),z) centered along the dipole (at z = z') and its image. At 3 GHz, 1 cmis equal to A/10; thus, at ESD frequencies like those shown in figure 6, thespark gap (which corresponds to the dipole current source at z = z') will beelectrically short and a dipole model is reasonable. The fields may beanalyzed using time -dependent scalar and vector potentials. The details areleft to Appendix A.
Although figure 7 shows the dipole above ground, actual ESD events takeplace very near the victim conductor. Thus, we will let the dipole approachthe ground, that is, z' =0. With this condition the electric field is
ound to be (see eq. A23)
-,- , - ^0 pz ,3 . 1 3i
R R
n
where R and r are defined in figure 7, and i is the current waveform as a
function of time evaluated at t - R/c . If the current waveform is known,this expression can be used to examine the resultant electric field eitherat the ground plane, which represents direct penetration into susceptibleequipment, or in some spatial direction as in the case of radiation to a
nearby victim. There are two basic factors controlling the magnitude of E.
2First, geometry factors such as l/(cR) and 1/(R ), and second, the currentshape, including its derivative (i, 3i/9t). For example, if we considergeometry as the dominant factor, then in the near field (very small R) E
should behave according to
E(r,t) ~ dil -^ ia ^ + a [^ - l]} ^ i (near field). (3)
Equation (3) suggests that the near-field E is controlled by the currentwaveform directly. If the victim is sensitive to electric field peaks thenthe hand- forearm portion of the discharge is the most hazardous. If thevictim is upset by prolonged exposure, then the full-body discharge orcurrent tail will be the most contributing factor.
The far-field E (large R) is dominated by
i(^,0-«^|a^£|.ij^'-l]|i^|i (far field). WR R
Thus, here it is the derivative of the current shape that contributes most.This suggests that for longer interference paths, such as indirect radiationto nearby susceptible equipment, low-voltage sparks with fast rise times maybe the most damaging. However, if we again examine (2), we see thatvariations in the current time dependence could be more important than thegeometry factors. In any case, exact results can be computed by includingall the terms in (2) with a known analytical waveform i(t).
In reality, it seems unreasonable to expect that current waveforms canbe written exactly in terms of simple analytical functions and substitutedinto the field expressions. Thus, some approximation will be necessary. Asimple approach is to express a measured current waveform as a sequence oflinear splines, as outlined in Appendix A (eqs. A25-A30). This approach is
compatible with our measurement system since we acquire the currentwaveforms using a digitizing oscilloscope, and the data points may be useddirectly to generate the linear spline. This method should provide accurateresults, although there are some minor difficulties such as discontinuitiesin the derivative of the current waveform. An alternative is to approximatea measured current waveform as a summation of damped exponentials (see eq.
A-31) . Because the current is basically due to the discharge of RLCcircuits, this approach should prove reasonable. In fact, the approximatedexpression may be realizable by a combination of simple circuits of lumpedparameters
.
In addition to the time behavior of the electric field we may take the
Fourier transform and determine its spectral behavior. This again requires
that we know the current waveform. Expressions based on our linear currentand damped exponential approximations are developed in Appendix B. Thesemay be used to identify the dominant frequency components excited by an ESDsignature
.
3.2 Magnetic Field
Magnetic fields can be a greater interference hazard than electricfields because they penetrate low- impedance shields more effectively. Theyalso excite apertures and seams as do electric fields. They are bestreceived by low- impedance antennas, especially circuit loops. Thus,avoiding loops poses a difficult design problem because loops are oftendifficult to recognize. Boxleitner [5] discusses methods for minimizingloop area by proper routing techniques
.
The magnetic field due to ESD radiation may be analyzed as was theelectric field. We find that (see eq. A23)
K
In the near-field region, H should behave according to
H(r,t) ~ a^ dl ]- ^ ~ i (near field), (6)<i>
^TT Rpj2
while in the far field we find
"<''•" -%«R^i (far field). (7)
The near- field magnetic field depends directly on the ESD current shape asdid the electric field. Thus, we would expect the highest fields to beassociated with the highest current levels. The far-field magnetic field is
dependent on the time derivative of the current, as was the case with theelectric field. Therefore, the low-voltage, fast rise-time sparks cause themost interference to nearby equipment.
4. MEASURED ESD CUBIRENT AND ELECTRIC -FIELD WAVEFORMS
The measurement of ESD currents and the associated electric fields is a
difficult task because of the very fast, broadband signals involved. Thetime-domain instrumentation must respond to these signals with minimaldistortion and drift. The measurement system assembled here represents a
best effort using equipment available.
4 . 1 Instrumentation
The experimental setup used to measure the ESD characteristics is shownin figure 8. It consists of three separate parts: the source, the target,and the measurement system.
8
1
The source was a commercially available ESD simulator indicated in fig.
8a and was made of lumped -constant circuit elements to simulate a human bodycharged to a preset voltage level. The source is capable of generatingvoltages from 100 V to 25 kV. It can be set to produce successive discrete
sparks at a chosen time interval.
One of the important targets we used in the experiment was provided bythe same manufacturer who supplied the ESD source. It consists of a brassball 8 mm in diameter (see fig. 8a), which is connected to a ground planethrough a parallel cluster of five low-value carbon resistors. It thus
provides a low-resistance path between the source and ground and a method of
determining the ESD spark current by measuring the voltage across the
resistors, since this voltage is proportional to the current through the
resistors. This voltage was measured through an extended center conductorof a coaxial cable which was connected to the ball target through a matching50-0 resistor. In fact, this is the only target with which we can measurethe ESD current.
The field measurement system consists of a newly designed broadband,UHF (NBS 200-1), time-domain TEM horn antenna (see Appendix C) placed over a
conductive ground plane, a digitizing oscilloscope, and a desktop computer.The digitizing oscilloscope has two inputs enabling us to measure thecurrent and radiated electric field simultaneously, as indicated by twodifferent colors. The system was calibrated by measuring the impulseresponse of the individual components (for example, pads, coaxial cable,connectors, oscilloscope) and then deconvolving out these componentfrequency responses to give the true current and field waveforms.
The antenna has a flat frequency response (± 3 dB) from 40 MHz to 1.2GHz with minimum phase distortion, which makes it ideal for measuring thebroadband signals in the time domain. The ground plane is an indoorfacility which has a finite time window. This time window is estimated bythe indirect scattering fields interfering with the direct signal betweenthe source and the antenna. The shortest reflective path determines thewidth of the time window. The facility used to make these measurements hasa direct path which is approximately three meters shorter than the smallestreflection path, thus allowing roughly a 10 -ns window. It is thereforepossible to measure the spectral content of fields at 100 MHz and greater.
The oscilloscope used in this experiment has a sample rate of 40
megasamples per second, which is not sufficient to capture the high-speedpulses generated by a single ESD spark. It was therefore necessary to
average successive ESD events to utilize the available bandwidth of 1 GHz.
The measurements were deemed repeatable enough to allow this averagingwithout compromising the integrity of the signals. Measurements of ESDcurrent and electric field were also made with a microwave antenna (NBS 50-
2), which is flat (± 6 dB) with even broader frequency coverage (than the
UHF, NBS 200-1) from 50 MHz to 4.5 GHz and a digitizing oscilloscope with a
bandwidth of 18 GHz in order to verify that there was a negligible spectralcontent at frequencies beyond the range of the oscilloscope and UHF antennaused in the original measurements. No discernable decrease in rise and fall
times was observed. Therefore, no further experiments were performed withthis system.
Basic functions of the desktop computer are for data acquisition,digital deconvolution, and digital filtering.
The most likely cause of error in the voltage (thus current)measurements (fig. 8a) is the assumption that the ball target consists of a
pure resistance over the entire frequency range of interest up to 1 GHz. It
is possible, however, that the carbon resistors are somewhat reactive forcertain frequencies causing the target to act as a low-pass filter.
The most likely cause of error in the field measurements (fig. 8b) is
the lack of isolation of the ESD spark. As a result, the antenna receivesfield compnents not only from the discharge to the ball target but also fromthe ESD simulator. In fact, slight changes in the simulator positioningcaused drastic variations in the resulting measured electric field. Forthis reason, every attempt was made to shield the simulator from themeasuring antenna by using absorbing material. With the absorbers in place,the measured fields did not change significantly when the simulator wasrepositioned. The discharge to the ball target was then the primaryradiation source as desired.
4.2 Experimental Results
Experiments were performed for a number of different targetconfigurations. The targets used were (1) the low- impedance ball mounted ina ground plane (fig. 8a), (2) direct sparking to the ground plane, (3) a
vertical, square (0.91 m on a side), metal plate mounted above a groundplane, (4) a metal chair insulated from the ground plane, and (5) a metalwaste basket. Targets (2) through (5) may simulate some practical officeenvironment. The low- impedance ball target allows us to capture both thecurrent waveform and the transient electric field simultaneously. In allother cases only the electric field was measured. We also recorded the
relative humidity of the room, the spark gap, and the minimum voltage at
which we were able to generate a spark by the simulator. Typically, the gapwas adjusted to create a preselected level for the minimum firing voltage.Thus, we quote only the firing voltage in the data discussion. In our
measurements, we use only a stationary gap because our main objective was to
vertify the theoretical model outlined in Section 3. Approaching a
simulator (ESD source) toward a testing object with a variety of fasterspeeds such as that which was studied recently [12-14] would have increasedthe leading edge slope (rise time) of the spark current, and accordingly,the radiated field. Thus, in this sense the experimental data presented in
this report are considered conservative.
Seven voltage levels up to 12 kV (1, 2, 4, 6, 8, 10, and 12 kV) wereset to excite the coaxial ball target. For higher voltages, the cleanwindow time on our indoor ground plane is not sufficient to obtainmeaningful fields data. The measurement arrangement is shown in figure 9.
The antenna was located at a distance of 1.5 m from the target at an angleof 20 with respect to the ground plane. At this distance the antenna willbe most likely in the far-field range (distance no less than A/27r, where A
is the wavelength) of the ESD source for frequencies above 30 MHz. As
mentioned before, the ESD simulator itself also radiated significantly. To
reduce this effect, absorber was placed between the simulator and the
antenna. The antenna was positioned to receive the vertical component of
10
the electric field. The horizontal component is expected to be smallbecause both the spark and target are vertically oriented [it can also be
seen from the first term in (2)]. Measurements of the horizontal electricfield confirmed this.
The results are shown in figures 10 through 16. Each figure shows the
measured current, the measured vertical electric field, and the predictedvertical electric field based on (2) and (A29) . For the 1-kV discharge, the
current reaches a peak amplitude of 7 A with a rise time of 700 ps. Themeasured vertical electric field reaches a peak of more than 50 V/m in about350 ps (approximately half the rise time of the current pulse) . As
expected from (4) , the electric field behaves very much as the derivative of
the current waveform. The predicted (theory) vertical electric field is
qualitatively similar to the measured data with two exceptions: first, the
peak amplitude is not as high as measured and second, the measuredundershoot is significantly greater. The first difference is notunexpected. In order to apply the theoretical model we need to know the
length of the radiating dipole (di) . We simply took it to be the height ofthe ball -target above the ground plane. However, the target consists not of
a single vertical conductor, but rather of an array of parallel resistorswhich act somewhat like an array of lossy vertical dipoles. Thus, it is notobvious how to specify an equivalent dipole length in this case. Inaddition, the tip of the simulator is a small loop which may itself beradiating and is also unaccounted for in the model. The second difference(undershoot) is more problematic. It could be an antenna response artifactbut in that case the system deconvolution should take it out. It may alsobe that the target is acting somewhat as a low-pass (capacitive) filter dueto the high-frequency response of the parallel resistors. The targetresponse is not accounted for in the deconvolution. The reason that thissuggests itself is the expected derivative behavior of the field response.The negative slope of the current waveform in figure 10 is not fast enoughto cause the observed undershoot. If the target response is slowing theleading and trailing edges, then both the peak amplitude and the undershootwill be underestimated by our theoretical model, as we are seeing. A targetwith ideal, flatter frequency response should give better agreement betweentheory and experiment. Nonetheless, the theoretical model presented heredoes predict well the basic electric -field behavior. The magnetic field mayalso be computed by (A29) with the same approximated current based on themeasured result. In particular, under the ideal far- field condition, theamplitude of the computed magnetic field in A/m differs approximately fromthe electric field by 1207r. Because no broadband H- field sensor is
available for measuring the magnetic field to compare with the theoreticalresult, only one computed magnetic field for the 4-kV case is presented in
figure 18
.
Figure 11 shows the results for a 2-kV spark. The current amplitude is
on the order of 11 A while the measured peak electric field is approximately75 V/m. Both are increased from the corresponding 1-kV values. Beyondthat, the same basic comments apply. Increasing the discharge voltage to 4
kV, as shown in figure 12, furthers this trend (26 A for current peak and160 V/m for electric field peak). The peak field is now quite high. We
expect much higher fields nearer the spark as will be discussed later.Raising the discharge voltage to 6 kV now reverses the growth trend. Thepeak current has fallen to 21 A while the peak electric field is down
n
significantly to 60 V/m, as shovm in figure 13. For the 8-kV spark shown infigure 14, the current amplitude remains about the same (approximately 24
A) , but the rise time of the waveform continues to slow. This reduces theradiated electric field. Similar comments apply to figure 15 (the 10-kVdischarge) and to figure 16 (the 12-kV discharge) . For the latter case thepeak electric field is down to just 20 V/m. Thus, the field interferencethreat appears to be associated primarily with lower-voltage sparks. Here,the 4-kV case appears to radiate most strongly. The reason for this wasdiscussed in Section 2 . As the spark gap increases and the dischargevoltage becomes higher, the channel formation slows and the channelresistance increases. These tend to slow the discharge current. However,the discharge now contains more energy due to the increased potential.
As the rise time of current waveforms becomes slower when the dischargevoltage is increasing, the theoretical predictions become better. Infigures 15 and 16, the deconvolution process of these slower pulses hasadded some noise to the current measurements. This appears clearly in the
theoretical curves, but qualitatively, the agreement between theory andmeasurement is good. This supports the supposition that it is the targetresponse to very fast currents that is causing the disagreement at the lowerdischarge voltages
.
We were not able to place our broadband receiving antenna closer to theESD spark since the antenna calibration required that it be in the far- fieldzone. We can, however, use the model to predict the discharge near fieldaccording to (2) or (3). Figure 17 shows the vertical electric field forthe 4-kV (worst case) spark of figure 12. The distances chosen were 1.5mas before, plus 1 m, 50 cm, 20 cm, and 10 cm. The near-field to far-fieldtransition frequencies for these distances are respectively 32, 48, 95, 239,
and 477 MHz. Thus, at 1.5m, most of the received spectra would be farfield while at 10 cm, a good portion is in the near field. The electricfield increases as the observation point is brought nearer until at 10 cm a
peak amplitude of over 4 kV/m is encountered. The waveform shape also tendstoward the current profile (rather than the time derivative of the currentfor the far-field case) as indicated by (3). The field levels at 10-cm, 20-
cm, and 50-cm distances in figure 17 are strictly theoretical and need to beconfirmed through measurements . But they suggest that the fields due to
low- level discharge voltages can be extremely high in the vicinity of the
spark.
The ESD electric-field spectrum for the 4-kV spark of figure 12 is
shown in figure 19. The primary spectral components are below 1 GHz whichis consistent with the findings of Honda and Ogura [1]. Above 3 GHz, the
spectrum is down to the noise of the system deconvolution. We have also
observed field pulses faster than the 4-kV case discussed here. They are
difficult to capture and analyze because of limitations in our digitizingoscilloscope. But we expect that the spectral content of some sparks may be
quite strong up to frequencies on the order of 5 GHz.
The preceding results were for discharges into a coaxial target which
allowed us to capture both the current and the field. However, as
discussed, the target may be filtering the current somewhat and it certainly
represents an idealized victim. Figures 20 through 23 show measured data
for the vertical electric field created by a spark to a ground plane at
12
discharge voltages of 1, 2, 4, and 6 kV respectively. The antenna waspositioned as in the ball- target measurements. The trend is similar to thatabove; the peak field level grows from 80 V/m at 1 kV to 100 V/m at 2 kV andthen begins to fall to 75 V/m at 4 kV and 45 V/m at 6 kV. The maximum peakis smaller than ball- target discharges. But in this case the radiatingdipole is significantly shorter. The largest gap was 1 . 2 mm for the 6-kVcase as compared to the 39 -mm spark gap for the ball -target measurements.Referring to (2) , we expect that reducing the dipole length di by an orderof magnitude would reduce the fields by a similar amount if the currentwaveform remained the same. However, measured data suggest that the fieldsare actually higher for the 1 and 2-kV cases (figs. 20 and 21 vs. figs. 10
and 11) and slightly reduced for the 4 and 6-kV cases (figs. 22 and 23 vs.
figs. 12 and 13). The ground plane is a true low- impedance target with a
very fast response (as opposed to the coaxial ball target necessary for
current measurements) and should not slow the current discharge. Theseobservations again imply that the actual current waveforms may be far fasterthan we have been able to measure here with the ball target. Regardless,the field levels for these discharges are considerable.
Perhaps more realistic than coaxial targets or ground planes are sparksto large metal objects, such as might be encountered in a normal workingenvironment. As an example, we consider the indirect radiation from a
large, square, vertical, metallic plate (0.915 m per side) as shown in
figure 24. The plate was excited by a discharge to the back side so thatdirect radiation from both the spark and the simulator are shielded from theantenna (again positioned 1.5 m away from the plate). A high- impedancestrap from the plate to the ground plane was also used to bleed off theacuumulated static charge, but this should not contribute greatly to theradiated field. The measured data for 5 and 7-kV sparks are shown infigures 25 and 26. Rather than receiving a single pulse as in the previouscases, the plate tends to resonate after being excited by the spark. Thefields are higher for the 5-kV spark, with the peak amplitude reaching 100V/m. There is definitely some contamination of the data due to reflectionsfrom walls and other scatterers after the 10-ns clean window time has beenexceeded. However, the contamination is not believed very significant dueto the highly directive property of the antenna. At least, the peak valueof 45 V/m at 14 ns is quite real. Even when the contamination is
substantial, the case may represent a more realistic laboratory environment.The 7-kV spark plate radiation is not nearly as significant. Only the
shadow- side radiation was tested here. A thorough study by including the
antenna- side exictations , edge excitations, variations in plate size, andother excitaion voltages would be of future interest.
Figure 27 shows the measurement configuration for radiation from a
metal chair above the ground plane. The chair was insulated from the groundplane by plastic tabs on its feet and was also provided a high- impedancepath to bleed off the static charge as for the plate. The antenna remains at
a distance of 1.5 m away from the spark point as before. The chair wasexcited by 3 and 6-kV sparks. The results are shown in figures 28 and 29.
Like the plate, the chair tends to resonate. For the 3-kV spark excitationthe field amplitude is quite substantial. Again, some of the data may be
contaminated after the 10-ns clean time. However, the basic indication is
that the indirect field radiation is significant and long lived. The 6-kV
spark excites a weaker field.
13
The measurement for a metal trash can in direct contact with the groundplane is shown in figure 30. The antenna is also placed 1.5 m away from thespark point. The received vertical electric field due to a 4-kV spark is
shown in figure 31. The initial spike within the system's clean timeexceeds 100 V/m, again demonstrating the threat of indirect radiated fields.
5. CONCLUDING REMARKS AND DISCUSSIONS
We have theoretically modeled the radiation from electrostaticdischarges by a simple elementary dipole above a ground plane. Radiatedfields were found to be dependent on two factors: the magnitude of the
transient current and its rise time. One factor may be more dominating thanthe other, depending on whether the observation point is in the near- fieldor far- field region of the ESD spark. When an analytical expression for the
ESD spark current is known, the electric and magnetic fields radiated fromthe ESD can be computed exactly by the theoretical model. We alsoimplemented an experimental system with a newly designed broadband antennaand a time-domain facility to measure both the current (for one case with a
ball target) and radiated electric field waveforms produced by a stationaryESD simulator. We approximated the measured current waveform with ananalytical expression, computed the corresponding electric and magneticfields , and found that the theoretical electric field agreed well with themeasured result. The electric fields due to ESD events could be quite high(although brief) , which are certainly capable of upsetting a nearbyequipment. The radiated magnetic field was not meausred because of non-existence of a broadband H- sensor.
If better experimental results are to be expected, several changescould be made to the system. The better ball target with a constantimpedance over the entire frequency range of interest and having a lowprofile and minimum radiation could be used to yield a more accuratemeasured current. A better shield for the ESD simulator is also needed so
that the source will not interfere with the measurements
.
It would be interesting to perform field measurements at positionscloser to the discharge to see if the transition from di/dt to i , inaddition to the change in peak amplitude, may also be observed. Moreaccurate measurements could be realized by using an outdoor ground screen to
obtain a much larger time window, thereby including the low- frequencycontent of the radiated field.
Measurements of magnetic fields constitute another interesting researchtopic, particularly in view of the fact that most shields are ineffectiveagainst low- impedance fields. Until a broadband H- antenna becomes a
reality, we can only rely on theoretical predictions and estimation derivedfrom the measured electric fields.
14
6. ACKNOWLEDGMENTS
We would like to thank Peter Richman, President of KeyTek InstrumentCorp., who inspired us to set up this modest study and convinced us that the
results would benefit the computer industry. Mr. Richman also criticallyreviewed this manuscript. We would also like to express our appreciation to
our Center Director, J. C. French, for his support and encouragement. Theenthusiastic assistance by our technician, H. W. Medley, during the courseof measurements is also acknowledged.
7. REFERENCES
[1] Honda, M. ; Ogura , Y. Electrostatic spark discharges - three criticalfactors. Proc. EOS/ESD Symp
., EOS-7, 149-154, 1985; Minneapolis, MN.
[2] Richman, P. ESD simulation - configuring a full -performancefacility. IEEE Int. Symp. on EMC; 381-385, Aug. 1983; Crystal City,VA.
[3] Bhar, T. B. Electrostatic discharge failure mechanisms ofsemiconductor devices. ITEM, 276-282, 1986.
[4] Byrne, W. W. The meaning of electrostatic discharge (ESD) inrelation to the human body characteristics and electronic equipment.IEEE Int. Symp. on EMC; 369-380, Aug. 1983; Washington, D. C.
[5] Boxleitner, W. Electrostatic discharge. Key Tronic Corp. Doc. No.36-2464, 1986.
[6] Woods, M. H. ; Gear, G. A new electrostatic discharge failure mode.IEEE Trans. Electron Devices, ED-26, 16-21; Jan. 1979.
[7] Penning, F. M. Electrical Discharges in Gases. New York, NY:
McMillan Co. ; 1957.
[8] Rhoades , W. T. Achieving ESD equipment protection with emissioncontrols. IEEE Int. Symp. on EMC; 232-237, Aug. 1985; Wakefield, MA.
[9] lEC Pub 801-2: Electrostatic Discharge for Industrial ProcessControl
.
[10] Richman, P; Tasker, A. ESD Testing: the interface between simulatorand equipment under test. 6th Int. Electromagn. Compat . Symp. Rec.;25-30, March 1985; Zurich, Switzerland.
[11] Richman, P. Classification of ESD hand/metal current waves versusapproach speed, voltage, electrode geometry, and humidity. IEEE Int.
Symp. on EMC; 451-460, Sept. 1986; San Diego, CA.
[12] Daout , B.; Ryser, H. Fast discharge mode in ESD-testing. 6th Symp.and Tech. Exhibition on EMC; 41-46, March, 1985; Zurich, Switzerland.
15
13] Daout, B; Ryser, H.; Germond, A.; Zweiacker, P.; The correlation ofrising slope and speed of approach in ESD tests. 7th Symp . and Tech.Exhibition on EMC; 461-466, March, 1987; Zurich, Switzerland.
14] Daout, B; Ryser, H.; The reproducibility of the rising slope in ESDtesting. IEEE Symp. on EMC; 467-474, Sept., 1986; San Diego, CA.
15] Richman, P. Computer modeling the effects of oscilloscope bandwidthon ESD waveforms, including arc oscillations. IEEE Int. Symp. onEMC; 238-245, Aug. 1985; Wakefield, MA.
16] Le Vine, D. M. ; Meneghini , R. Electromagnetic fields radiated from a
lightning return stroke: application of an exact solution to
Maxwell's equations. J. Geophys . Res., 83, no. C5 ; 2377-2384, May1978.
17] Lawton, R. A.; Ondrejka, A. R. Antennas and the associated timedomain range for the measurement of impulsive fields. NBS Tech Note1008, Nov., 1978; Boulder, CO.
16
APPENDIX A: Time -Dependent Fields Due to a Dlpole Source
The fields due to a time-dependent radiation source may be found from the
potentials <I>(r,t) and A(r,t) which are given by
^r ^^ 1 f?(•"' ,t-R/c)
, ,$(r,t) =J.— / ^ 5 dv', and
(A1)
A/- .^ ^0r J(^' ,t-R/c)^
,^^^'"^ = H^ ^ -R^"
•
where r is the observation point, r' is the source point, R = |r - r' | is the
separation between the two, p is the charge distribution, J is the currentdistribution, c is the speed of light, e^^ is the free-space permittivity, and
]x is the free-space permeability. The fields are related to the potentials
according to
E(r,t) = -V$(r,t) - |^A(r,t), andat
H(r,t) = - VxA(r,t).^0
The potentials satisfy the condition
V-A(r,t) + ^ It'l'^^'t)= 0,
c
which implies that
t
/
(A2)
<I)(r,t) = -c^ / V-A(r,t')dt' . (A3)
Thus we need only solve for A(r,t).
Consider a time-dependent dipole source above a perfect ground as shown in
figure 7. We impose a cylindrical coordinate system and note that the systemis invariant in ((). The fields for z ^ will be due to the dipole source and
its image. The current distribution J reduces to an elemental line current i
oriented along the z-axis, the volume integrals simply introduce the linelength dS,, and (A1) reduces to
A(r,t) ="l
A.(r,t), where
(Ai4)
17
u i(z' ,t-R ./c)
A.(r,t) = a dl ^— .
J z 4iT R
.
In (AM) j = 1 represents the actual dipole and j=2 its image. We may now solve
for $(r,t) . We find that
y 2 i(z',t-R /c)
V.A(r.t) = dd ^ l^[ R—^^ )• (A5)
j = 1 J
If we let u = t - R./c, and let the argument of i be implicit, then
|_(i
) = h:2 {R. |i|H - i_J|.
. (A6)9z R. J
^
J 3u 9z dz '
Note that
3i 9i
3u 9t
1 9R.9u 1 J
(A7)
9z c 9z•
Thus
,
i(z',t-R /c) 9R R
I- [ 5^
]= - R:^ ^^ {i - -^ ^], (A8)
9z ^ R . ^ J 9z ^ c 9t
'
and
M 2 _ 9R.
^- ::>
V-Mr.t) .-«^^z^r:2_J |i . J|il. (A9)
It follows that
c^y 2 9R . t R
We may now solve for the field components.
We will begin with H(r,t). Substituting (A4) into (A2) yields
V«A(r,t) . 61"t^
Z V«(a^ i) - ^ £ dl ^ £ |
(i). (AH)
The p-derivative is analogous to the z-derivative found above. Thus,
1
U 2 8R. R
VxA(r,t) = a dil -^ Z r/ ^-^ (i + -^ ^M . (A12)(}) 47T ._ J dp C dt'
Thus we have
p R
S(r.t) = a^dJ I^ ^Z^R-2 £^ (i .J ii). (A,3)
In order to find E(r,t) we need to evaluate V$, where
V$(r,t) = a ^ <J)(r,t) + a | $(r,t). (AU)p dp Z dZ
Taking the proper derivatives yields
^P J ^^ ° ^^ R^ R^ cr2 A.^^J J J J
(A15)
- „9R.t R... . _,-,,2t3z J 3z Q c 3t p3 p2 Q
J J
1 r^ _ 3(z+z')^^. _ 1 (z+z')^3i
cr2^
r2'
e^R. r29t'
J J J J
where + designates the source (-) and its image (+). Thus we find that (note2
that c y^ = I/Eq)
VMr.t)=-if- z^M^I^-^i ^4-lfiP "^^^0
j = l r2 r3 cr2 c^R.StJ J J J
(A16)
,- dJl I \.
3(z+z' )^, ^ 1 r. 3(z^2' )^. 1 (z + z')^8i,
J J J J J J
where
t
I = I(z',t-R./e) = /i(t'-R./c)dt' .
J Q J
Next we calculate
3A(r,t) ; ^„ ^0 ^ 1 3i - .. 1 ^ 1 3i ,..„.—r- = a dJl -J— E 5" -^ = a dil j— Z -^ x— . (A17)
dt z 4tt . , R. dt z ^Tr£_ . , 2^ dtj = 1 J j = l c R
19
Combining results we find that
E(r t) = a ^ E|P ^"^"'^
[3 ^ , _i_. , _L ill l
Oj=1j J R. cR. cR.J J J
(A18)
J J J J J
For a discharge to a conductor the spark will occur very close to the+
conductor surface, that is z' = . If we impose this condition we find that
R. = R = (z + p ]and that
-,- ^, - di pz ,3 ^ 3 . 1 3iiE(r,t) = a -— ^ — I + —-1 + -— 3-
P 2^^0 r2 r3 cr2 c^r9^
2 26i f r 3z , -1 /- 1 .^ 1 . ( rZ
, 1 1 3i 1
(A19)-,- ^v - dil p r1 . 1 3iiH(r,t) = a :r- ^ - 1 + — —
.
(j) 2-iT R '2 cR dtK
As a check note that in the far field the 1 /R terms should dominate. If
we convert to spherical coordinates we find that
^,- ^v - dS, sine 3i
e 2lT£_ ^r^ dtc R
(A20)
dil sine 3i
2^ cR 3t'H(r,t) - a
Thus the wave impedance tends toward n^, as expected.
Although, the above expressions have the correct radiation behavior, thereis difficulty with this model in the very near field. Consider a static
electric dipole P of length dil along the z-axis,
P = a dilQ, (A21 )
z
where Q is the total charge on either pole. Note that Q is a measure of the
net current either leaving or entering the pole,
t
Q = ± / i(t' )dt',
(A22)
20
± depending on which pole is being considered. Thus, Q is related to I. The
3appearance of a 1/R , static electric-dipole moment contribution is a result of
2the assumption that the dipole is short (dSL/R << 1, k (dil) /R << 1). If
instead we had considered a linear antenna with appreciable length and assumedthe current to be a travelling wave, then the static I-dependent dipole momentterm does not arise [16]. Although a dipole in free space can hold a staticcharge, our ESD system will not. The charge will actually decay rapidly due tothe relatively low relaxation time e/o on a good conductor. Thus, we willneglect the I terms in our model. The field expressions therefore reduce to
^r^ ^^ ' ^0 ^0 pz r3 . 1 3iiE(r,t) = a dil :r- -^ Pr 1 + — ^T-
p Z-n ^2 '2 cR 9tn n
^a^«^l[2|'-,]i5iM^J-l]isil, and (A23)
R R R
r, r~ . \ ~ ,„1 Pfl . 1 3iiH(r,t) .^Al-l\-..--\.
R
Equation (A23) allows us to consider the fields due to a short dipole neara ground plane. We wish to model either direct field penetration into thevictim in which case z = 0, or the effect of radiated fields on nearby objects,for example due to a discharge to a metal cart. For the z = case (direct
victim penetration), R = p, r depends only on p,
^0 (1 . 1 9iE(p,t) = - i dSl -^ {^ i + ;^ fi}, and
z 27r 2 cp dtP
S(r,t) = a dJi ^ |1- i . i-|id) 2tt 2 Cp dtY p K
(A2i4)
Thus, on the ground screen in the transitional near field, not too close to thedipole as per our assumptions, the wave impedance retains the far-field valueof 120iT Q.
In general the time behavior of an actual ESD current will be quiteerratic; thus, we do not expect to be able to model it by a simple analyticfunction. One approach is to approximate the current by a series of linearfunctions, each valid over a small interval. By choosing the number ofintervals to be large we should be able to generate an accurate model. Inparticular, we will use a set of ramp functions
N
I
n=0i(0^,t) = I i^(t), (A25)
where
i (t) = for t S t ,
n n
21
At
i (t) = a^ —n n A
for t < t < t , , andn n+1
(A26)
i (t) = an n
for t , S tn+1
where At
see that
t - t , and A = t ,- t ,
n n n+1 nWith this linear approximation we
S + ^3-|^i(0 ,t-R/c) = I f,i^(t-R/c)at ^ dt n
n=0(A27)
Evaluating (A23) yields
^^i =8t n
for t-R/c S t ,
n
3 . ^ _3t^n A
n
n
for t S t-R/c St , , andn n+1
(A28)
at^n = ° for t , < t-R/c.n+1
Thus, for this approximation we have a discontinuous derivative. If wesubstitute these into the field expressions (A23) we find the followingapproximate forms
E(r,t) = dS, ^ la ^ Z E + a E E } , and2tt ' p ^2 ^ - pn z „ zn"
R n=0
.„ N
H(r,t) = a fjf I H<}) 2iT R ^^Q (t)n
n=0
(A29)
where
E =pn
n 1 RiE = -^ ^ |3At + -}pn A „2 ' n c'^ n R
E -a 2^pn n ?
for t - R/C < t ,
n
for t < t-R/c S tn
for t , < t-R/c,n+1
n+1'
E =zn
for t-R/c < t ,
n(A30)
E = -H 1^ {(1| - i]At +[^-i)^} fort < t-R/c < t ,.zn A p,2 ' ^^2 ^ n *-^2 ^c' n n+1
n R R R
E = a i. [%- - 1],zn n ^2 \2
for t , < t-R/c,n+1
H =(j)n
for t-R/c S t ,
n
22
a
H, = T^{-^ At + —
}
for t < t-R/c < t , , andAn A ^^2 n cR
'
n n+1n R
H, = a ^ for t ,<t-R/c.4)n n ^2 n+1
K
These may be readily found via computer once a linear spline approximation to a
current waveform is given.
An alternative to a direct point-matching approach such as outlined above,is to fit a sequence of damped exponentials to the current waveform based onProny's method. Let
N -b t
i(t) = E a e " (A31
)
n=1 "
for tSO and i(t) = for t<0. Continuity thus implies that the summation ofthe a be zero. The derivatives of this form are readily found and the result
n •'
will be continuous. The derivative will not be continuous at t = unlessfurther conditions are imposed. Inserting the exponential fit (A31 ) into thefield expressions (A23) yields
T1-. N -b t _ .
i(r.t) = i dil ^£| Z a e " [\ - 1^ bjP ^TT ^2 ^^^ n ^^2 cR n^
n=
1
n R R
^ n=1 R
23
APPENDIX B: Spectrvim of ESD Dipole Radiation Model
If the field components are determined in the time domain, the spectrummay be found by taking the Fourier transform according to
00
F(r,u)) = / F(r,t)e'^'^^dt. (B1 )
where F represents one of the fields. Substituting in the field expressions
for E and H (eq. A23) leads to two integrals that need to be evaluated:
/ i(t) e'-^'^^dt, and
(B2)
, 9i(t) ^-ja)t
If the second integral is integrated by parts we find that
/lil^e-J^'dt = i(t)e-J'^^d t
^° + JO) ; i(t) e -^"^^dt. (B3)
For a pulse of finite duration, i(t) = at both the endpoints and (83) reduces
to
/ ^^ e'J'^'dt = jo) ; i(t)e"J'^'dt. (BM)t
-00 —00
Thus, we need only consider this single integral in order to evaluate (B1).
We will again represent the current by a sum of linear functions (A25) as
detailed in Appendix A. The pulse is zero outside the interval t e (0,t )
where we will evaluate the integral in terms of retarded time (t-R/c -> t).
Thus we have
^N+1 _ N ^N+1
/ i(t) e'J'^^dt = I ; i (t) e'-^'^^dt
n=0 "
(B5)
N a ^n+1 . ^ ^N+1 , .
= I {-p. S (t-t ) e'J'^^dt -. a ; e'J'^^dt}.
n=0 ^ t-^ " t
n n
Evaluating the various integrals and combining results yields
24
Vl . , N a "J'^^n -ja)A "^'"^+1 N
/ i(t) e'J'^^dt = Z -^ ^-^ je '' - 1 1 + j ^ Z a . (B6)
n=0 ^ 0)2^
n=0"
For a pulse of finite duration, the a summation is zero. For convenience, let
S(u») denote the remaining summation in (B6); that is
N a "J'^^n -j(oA
S(a)) = Z ^ ^-j— {e ^ - l}. (B7)
n=0 n 0)
The field component spectra are found to be
R n
'h^''^ n^^- 1)-^ - (4 - 0^15(03). and (B8)
R R R
H(r-..) -i. #1 li. * i|ls(a,).<f>
2Tr R '2 cRn
We may also consider the damped exponential form for the current given by
(A31). In this case the equivalent expression to (B5) is
/ i(t) e •J'^^dt = Z a / e " dt
n=0 "
(B9)
N1
= Z a^ . ' = T(w).n=0 " t)^ ^ JO)
The spectra of the E and H fields will be again given by (B8) with T ( oj
)
substituted for SCoj).
25
APPENDIX C: Time-Domain Antennas for Measuring Impulsive Electromagnetic Fields
Since the purpose of our experiment was to study the temporalcharacteristics of the electromagnetic fields emitted from ESD events, it wasimportant to choose a broadband receiving antenna that had the greatestpossible fidelity. Some antennas produced severe distortion in the time-domainwaveform because of their poor phase response even though they did have flatcharacteristics in the frequency domain. The antenna used for the major partof the work is called UHF antenna (NBS 200-1), which is basically a truncatedTEM horn as shown in figure 32. Two elements of this antenna form a parallel-plate transmission line, whose impedance depends on the ratio of the distancebetween the elements to their width. Since this ratio is constant, the antennaimpedance remains 200 Q within its designed operating frequency range (from 40
MHz to 1.2 GHz). At the antenna feed point, there is a pulse transformer, whichterminates, on one side, the elements in 200 Q and produces a 50 output onthe other side. The details of this pulse transformer are given in figure 33.
It is this pulse transformer which maintains the correct impedance up to 1.2GHz. Consequently, the standing waves of the antenna are negligibly small, andthe frequency response is flat within ± 3 dB in the designed frequency range of40 MHz to 1.2 GHz with usable response even down to 20 MHz, as shown in figure34. The time -domain response of this antenna to the transmitting pulse shownin figure 35 is presented in figure 36. This antenna has a risetime responseof about 300 ps
.
Another modified version called the microwave antenna (NBS 50-2) also hasexcellent fidelity and flat frequency response with a broader designedoperating frequncy range (from 50 MHz to 4.5 GHz). The antenna elements aredesigned to have an impedance of only 100 Q, each element of the antenna being50 Q with respect to a ground plane located in the middle of the antenna, as
shown in figure 37. Each element is brought out on a separate coaxial line.
The previously used pulse transformer is now replaced with a broadband balun(see fig. 38). The balun accepts two input signals which are 180 degrees outof phase with each other and combines them in a broadband resistive powercombiner (splitter in the figure) providing a single 50 Q output. Details ofthis antenna are available in [17]. This antenna was used only for that partof the experiment where we were searching potential spectra which may falloutside of the UHF antenna frequency range, as mentioned in Section. 4.1.
26
v////////. v////////y V//////A
b
v/y/y////
d
Figure 1. Diagrams showing the development of a spark channel, based
on cloud chamber observations: (a) electron avalanche, (b)
channel growth toward anode , (c) channel growth towardboth anode and cathode, and (d) conductive bridge [7].
27
RAAAr-
/DISCHARGE
TIP
RETURN\
GROUND PLANE
Figure 2. Conventional single RC circuit model for human electrostaticdischarge [9].
60-300
PF
.05 -.2
777
Figure 3. The dual RLC circuit model for human electrostatic dischargeincorporating separate, parallel paths for body and handdischarges [11].
28
Figure 4. Computer simulation of an BSD current based on the dual
RLC circuit (fig. 3). The circuit parameters used areC^=7.5 pF, R^=200 ti, L^=0.1 uH , Cg=100 pF, Rg=500 Q, Lg=1.7
yH, and 5 -kV initial charge fll].
29
8 12 16
t (ns)
(a) 4 GHz/0. 09 ns
Tp = 0.7 ns
4 8 12 16
t (ns)
(b) 1 GHz/0.35 ns
Tp = 0.7 ns
8 12 16
t (ns)
(c) 400 MHz/0.9 ns
Tp = 0.9 ns
18 4 8 12 16
t (ns)(d) 100 MHz/3.5 ns
Tp = 1.5 ns
t (ns)
(e) 75 MHz/4.7 ns
Tp = 1.7 ns
4 8
t (ns)
(f) 60 MHz/5.8 ns
Tp = 1.9 ns
Figure 5. Computer solutions for a dual RLC-circuit model using
different single-pole simulated oscilloscope bandwidths
The initial applied voltage is 5 kV [15].
30
V„ IkV Vo 3kV
-iODBI 1 0094Z
iii.t :hiK:. r1 t
200WZ/
Rinif'ipiiijifiiir^if mLiu:!,"
h—
r.' •
•
^'1001/ 200B a-J..8
1
• -
.
IMT
fai i
Vo 6kV V„ 10 kV
Vo 15 kV
{^1 gH^B^H-IODBI I 00li«Z
,
200W2/
Jilliii\\, !;.
* ^
. . -
... ••.
ijwrmMlIBjji£
Vo 20 kV
Figure 6. The radiation spectrum at 2.0 m from an ESD source for
various applied voltages [1].
31
r = (p ,(}),z)
Dipole z'I
Image Dipole
Figure 7. A dipole above a perfect ground plane and its image,
32
ESD
simulator
Ground Plane BallO
rarget i
DigitizingOscilloscope
10 dBPad
RG 55B CoaxCable 5.5 m
10 dBPad
(a) For measuring current waveform.
Reals tlvely-Terralna tedTapered-Horn
Antenna
DigitizingOscilloscope
10 dB
PadRC 55B CoaxCable ^.3 m
(b) For measuring electric field,
Figure 8. Measurement system schematics,
33
Figure 9. Coaxial ball target measurement configuration (Shield
material was used to prevent radiations of the ESD
simulator from interfering with the receiving system).
34
4->
CS-s-
30
25
20
15
10
-5
-10
-I r1 r
5_l I L.
Time (ns)
IKV SPRRK5 1 •/< Hum.2 Ohm TRRGET"
10 15
-a
•r—
U
O<U
LU
200
150
"T r r -\ 1 r
100t-
50 -
-50h-
- 100
Expe r i men t
IKV SPRRK5 1 *. hlum .
2 Ohm TRRGET
Time (ns)
Figure 10. Coaxial ball target current and vertical electric field
for a 1-kV discharge.
35
+->
ccuS-s-
o
30
25
1 1 < 1
' ' ' 1 2KV SPARK5 \y. Hum.2 Ohm TRRGET"
20 - -
15 - -
10/^~~-\ -
5
1
\^..__________1
-5 -
—
10. 1 1 . 1 1 1 1 . 1 1 1 ... 1 1
Time (ns)10 15
200
150
T 1 1 r
-o
03O
+->
O
LU
1001-
50
-501-
-100
2KV SPHRK5 1*^ Hum.2 Ohm TRRGtIT
Time (ns)
Figure 11. Coaxial ball target current and vertical electric fieldfor a 2-kV discharge.
36
cs-S-
C.I
30 1 1 1 1 1 1 1 I1
1 II T-
4KV SPRRK5 0>: Hum.
kfb /~\ 2 Ohm TARGET"
20 / \ -
15 / \ -
10/ \
5 - /^-^
-5
10
J
i .1 _ .L .. 1 , . . .
Time (ns)10 15
o•r—
M-
U•I
—
S--)->
uO)
200 1
150 -
100 -
50 -
-50 -
-100.i , I I 1 J-
Time (ns)
Figure 12. Coaxial ball target current and vertical electric fieldfor a 4-kV discharge.
37
4->
CO)S-s-
o
30
25
15
-5
10
'~l'
Time (ns)
T 1 r-
GKV SPRRK^SX Hum.2 Ohm TARGET"
10 15
aO)
u
-MOO)
200 [ 1111 II
' GKV' SPRRK4 9:: Hum.
150 - 2 Ohm TnRGET_
100 - -
1
1
Ex pe r 1 men t
50 -
y^=:^^V, \^ Theory ^^::;=^==~"===^3
-50 -^^^
-
1001 1 1 . . . . 1 . . . .
10 15
Time (ns)
Figure 13. Coaxial ball target current and vertical electric fieldfor a 6-kV discharge.
38
O)s-
o
30
-5 -
-10
-1 1 1 r -r r r
J 1 I
Time (ns)
BKV SPRRK^QX Hum.2 Ohm TRRGET'
-J I I u10 15
-a^—
^-
u•r—
O(D
UJ
200
150 -
100
-100
E X p e r 1 me n t
-1 I I u _i I I L-
Time (ns)
8KV SPRRK4 8:c Hum.2 Ohm THRGET
_L 1 L. J 1
10 15
Figure 14. Coaxial ball target current and vertical electric fieldfor a 8-kV discharge.
39
+->
S-s-
o
30 1 1
1
25 -
20 -
15 -
10 -
5 - /^Q
-5 -
-101 1
T3
oOJ
200
150
100
50 -
-50
- 1 00
-J I u
Time (ns)
10KV SPARK5 3 ^i Hu m .
2 Ohm THRGET"
-I 1 1 i_
10 15
lOKV SPF^RK5 3/. h^ u m .
2 Ohm TRRGET
Expe r I men t
I L I I L^ X I I 1 I I I i_
5 10 15
Time (ns)
Figure 15. Coaxial ball target current and vertical electric field
for a 10-kV discharge.
40
O)
=3C_)
30 J— . -, 1 -T- I I I I 'I ' " '' '
12KV SPRRKS5y, Hum.
2b 2 Ohm TRRGET"
20 ^ ^^-\_. "
15/^^'^ ^^^
10 /^5 '^^^^
-b
101 1 1 1 1 1 1 1 I I 1 1 1 1
Time (ns)10 15
200
150
^ 100
-a
^ 50u
-50
-100
-1 1 1 r
L'xper 1 ment
12KV SSPRRK5 5*. Hum.2 Ohm TRRGET
I 1 I
Time (ns)10
Th eo r y
J I \-
15
Figure 16. Coaxial ball target current and vertical electric field
for a 12-kV discharge,
41
•r—
+->
oOJ
5000
4000 -
3000
2000
1000
1000
-1 1 1 r
1 O cm
-I I I u
Time (ns)10 15
igu-e 17. Predicted vertical electric field for the 4-kV discharge of
fiq. 12 as a function of distance to the observation point.
42
r—dj•r™
OMCUcen
0. 4
0.3
0. 2
0. 1
0. 1
(a)
J I L J I 1 I J 1 1 U10 15
Time (ns)
20
15
10 -
E
-o
'r—4-
O+^ 5O)Ecn
-5
1 1 1 r T 1 1 r T 1 1 r
J 1 I L J I I L J 1 I L
10
Time (ns)
15
Figure 18, Predicted magnetic field for the 4-KV discharge case:(a) with the receiving antenna at 1.5 m from the spark, and(b) as a function of the distance to the observation point.
43
160
140
120;
1001-
> 80 -CQ
O) G0 -
13
.^J. ^
1. _._|_.^_j.- ^ I
CLE
40 -
20
0. I
Frequency (GHz)
Figure 19. Spectral content for the coaxial ball target 4-KV discharge of
fig. 12.
44
oO)
<^
uS-4->
UO)
LU
200
150
100
50 -
0-
-50
-100
I
1 .1—
IKV SPRRK50Ji Hum.
-I I I u -I I I L.
Time (ns)10 15
Figure 20- Vertical electric field for a 1-kV spark to a ground plane,
45
o•I—
s-
o
200 1 1 1 1 1 1 1 1 1 1 1 r 11
2KV SPRRK5Qy. Hum.
150 — -
100 A -
50
-50
j\ .-
^ ^-
1001. .. .. I . . 1 1 1 .1 1.1. I... —L 1 J 1— .1
10
Time (ns)
15
Figure 21. Vertical electric field for a 2-kV spark to a ground plane,
46
a"r—
M-
O•^"
S-•t->
oO)
UJ
200 1 1 1 1 . 1 1 1 1 1 1 1 1
4KV SPHRK50>i Hum.
150 — -
100 - -
50
-50
- A -
-
\y^-
-100 .ill 1 1 1 I 1 1 1 1
Time (ns)10 15
Figure 22 • Vertical electric field for a 4-kV spark to a ground plane,
47
-o
O)
oS-
u0)
200 1 . 1 1 1 1 1 1 1 1 1 1 1
6KV SPHRK50>i Hum.
150 - -
100 - -
50 -
/-X-
X^_/^ ^
-50 - -
1001 11 1 1 1 1 1 1 1 1 1 1 1
Time (ns)10 15
Fiqure 23- Vertical electric field for a 6-kV spark to a ground plane,
48
Figure 24. Measurement configuration for indirect radiation from a
vertical square metal plate.
49
150
100
-1 r— T r •—r-J-5KV
-T 1 T TSPRRK
irJTO 9 1 .5cmSQUHRE PLRTE
•r—
M-
US_4JO
-100 -
-150-I I I L -J 1 I U. J 1 1_. ' ' I 1-
10 20 30 40 50
Time (ns)
Figure 25. Vertical electric field radiated by a vertical square metal
plate excited by a S-kV spark.
50
150
100
^-^ 50-a
iT
u•I
—
s-+->
uOJ
|— -50
-100
-150
"T I T T 1 1 1 1 1 1 r -T 1 11
1 1 1 1
7KV SPRRKIMTO 9 1 . 5cmSQURRE PLRTE
.1 L I I I I I I I L 1_ -J I I L 1 I I I 1_
10 20 30
Time (ns)
40 50
Figure 26. Vertical electric field radiated by a vertical square metal
plate excited by a 7-kV spark.
51
Figure 27. Measurement configuration for radiation by a metal chairabove a ground plane.
52
150
100
T r—I1— 1
1 r—I
rV -1 r—-T-
3KV )PRRKINTO rJBS
OFFICE CHHIR
-o
%•I—
M-
O•p-
S-+Jocu
100
-15010
._! I I 1 i__L.-.
20 30
Time (ns)
40 50
Figure 28. Vertical electric field radiated by a metal chairabove a ground plane excited by a 3-kV spark.
53
-o
'a;
s-
o(U
150 1—11 1 1 1
1 1 1—11 1
—
\1 1—1 1 1 1 1 1 1
1BKV SPRRKIMTO NBS
100OFFICE CHRIR_
50 -
\
—^V^^^^^^^^
-50 -
100 -
1501 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
10 20
Time (ns)
30 40 50
Figure 29. Vertical electric field radiated by a metal chair
above a ground plane excited by a 6-kV spark.
54
Figure 30. Measurement configuration for radiation by a metal trashcan on a ground plane.
55
-a
U
150
100 -
50 -
-50 -
-100
-150
T1
1 1 1
4KV SPRRKINTO NBSTRHSH CRN
-I I I I-
10 20Time (ns)
30 ^Q 50
Figure 31. Vertical electric field radiated by a metaltrash can excited by a 4-kV spark.
Note:
Two, single-sided, GIO copper-cladglass epoxy PC boards, separatedwith a wedge of styroform extrudedpolystyrene insulation (copper sidedown)
.
Double-sided, GIO copper clad glass
epoxy PC board, with details shownin fig. 33.
Figure 32. The UHF antenna (NBS 200-1) for measuring the broadbandelectric fields: dimensions in the figure are in
centimeters.
57
17.73
NOTE 1
Note:
1. SMA connector for 0.36 cm semi-rigid cable.
2. 0.36 cm semi-rigid cable wrappedwith alternating ferrite cores of
Q1,Q2, and Q3 material types withtypical dimensions of 0.97 cmouter diameter, 0.74 cm innerdiameter, and 0.33 cm thickness.
3. PC board fits into 15.24 cm slotin foam wedge.
4. Single side copper clad antennapiece
5.08 H H
Detail A
0.32
/
GNO
FRONT
GNO
Figure 33. Construction details of the pulse transformer in the
NBS 200-1 antenna (all dimensions in centimeters).
58
owrj
u (U
m4-1 <c
3 oOx: j=
ooCM
mCO2:
(Ux: •
+-> -MCD
<4- J:^
UfO
cu •l->
toc X
O)Q.-a1/1 >,QJ 2s-
QJ>.-c
-McOJ +->
13 =3crOJ x:S- -MM- • r-
2OJ-C "O-(-> c
fO<4-
-C-t->
e •1
—
200
•1
—
fcs- efC EQ. O)E -M
CC_) fO
«
rr>
0)s-3en
oin
o O OfO CM
(ap) joqoBj euua^uy
59
Time (ns)
Figure 35. Normalized amplitude of the transmitting pulse,
1 —I 1 1
—
1
—
1 1
—
41
+ .8
o13+->
CLE(T3
DO)N
+.e
+ .4
^*\k
ES-o
•
-.2
-.4
-.G
3 12.^ 2 S.'fl 37.^ 55.5 -BIT^ 75.3 -TTTt- 99.
-.8
Time (ns)
Figure 36. Response of the NBS 200-1 antenna to thetransmitting pulse shown in fig. 35.
60
01
.-—
-—
\ '-^\ ^. ^w \ \
\
\ V\
\\\
\. \\\
.̂\\,
\
\ \^\ \ \\\\\\
,
N \ \\\\ \^\\^^VK \\\\
\
\\x\S
o\\vf \^^'\\
\\ \\^
\>1\ \\ V^\\ N\ \^\ ^ \<\\^
\\\ \. \ >-^\ \
\ \/
/c
\-o
/ /o/
/ / / / \^ / / /
/ / ///\ / / // / / 1/ A / / /
/ / y / ^
y / /
/ / / / / / \ / / /
/ / / ^ A / /
/ / /y/ / / \/ / /
/ //
/ / A / /
/ / / / / )•
/ /
/ // / / / \- / 1^
/ /IV / / /A / /
/ J / / / A' /
/ / / / / / \ / //
y1/ / / / / \ /
/ 7/ / / /
A '/ / / / / / / \/
/ h / / / / \•
/ / / / / / // / / / / / / /
u
t-l
(LI
X D.nj Q.
o
01
o.< 0*
O &0i-> c
CO
•O -HOJ u1- H<u
o -^^ 1-(
o oCO (b
uOJ 01
> -u(0 (0
oU QOu c•H tA
uo(0
Xi<
BaoCb
0)
c0)
u>«4JCO
>>iHO
OJ
oLn
CO
TOEc:
»a
ES-Ox:
>5oS-o
O)
on
13C7I
k e rj
o
^
61
10.0 cm.
B.nl.inced
Non-lnvertf(J l/ll ^^
10.5
cm
FerrlteBeads
1 ea. Ql and Q3
rm Thick \0.5
Aluminum
3 mm Scml-rlgldCoax
Outer ConductorsSoldered
.Space Killed Wltli
Microwave Absorber
6 dn Power
Splitter
R.I lanced
Inverted 1/(J PrecisionType N Connector
Inverting TSplice
.4=^
Unbalanced I/O. Precision Type N
Connector
.15.5 cm
Figure 38. The broadband balun used for the antenna in fig. 37,
62
NBS-n4A (REV. 2-BC)
U.S. DEPT. OF COMM.
BIBLIOGRAPHIC DATASHEET (See instructions)
1. PUBLICATION ORREPORT NO.
NBS/TN-1314
2. Performing Organ. Report No. 3. Publication Date
February 1988
4. TITLE AND SUBTITLE
ELECTROMAGNETIC FIELDS RADIATED FROM ELECTROSTATIC DISCHARGES-THEORY AND EXPERIMENT
5. AUTHOR(S)
P. F. Wilson, A. R. Ondrejka, M. T. Ma, and J. M. Ladbury
6. PERFORMING ORGANIZATION (If joint or other than NBS, see instructions)
NATIONAL BUREAU OF STANDARDSDEPARTMENT OF COMMERCEWASHINGTON, D.C. 20234
7. Contract/Grant No.
8. Type of Report & Period Covered
9. SPONSORING ORGANIZATION NAME AND COMPLETE ADDRESS (Street. City. State. ZIP)
10. SUPPLEMENTARY NOTES
I I
Document describes a computer program; SF-185, FIPS Software Summary, is attached.
11. ABSTRACT (A 200-word or less factual sun-innary of most significant information. If document includes a significantbibliography or literature survey, mention it here)
The fields radiated by electrostatic discharges (ESD)
and experimentally. The ESD spark is modeled theoreti
time dependent, linear dipole situated above an infinsparks of varying voltages are generated by a commercused to excite a number of targets including (1) thecoaxial cable mounted in a ground plane, (2) direct d
(3) indirect radiation from a large metal plate, (4)plane, and (5) a metal trash can. Results show that(2-4 kV) excite the strongest radiated fields. Thiscan pose a significant interference threat to electrogigahertz range.
are studied both theoreticallycally as an electrically short,
ite ground plane. Experimen-tally,
ially available simulator and
extended inner conductor of a
ischarges to a ground plane,
a metal chair over a groundrelatively low-voltage sparks
suggests that the spark fields
nic equipment into the
12. KEY WORDS (S/x to twelve entries; alphabetical order; capitalize only proper names; and separate key words by semicolons)
electrostatic discharge; pulsed current; radiated electric field; radiated
magnetic field; spectrum; time-domain technique
13. AVAILABILITY
[77] Unlimited
rn For Official Distribution. Do Not Release to NTIS
r~Xl Order From Superintendent of Documents, U.S. Government Printing Office, Washington, D.C20402.
[33] Order From National Technical Information Service (NTIS), Springfield, VA. 22161
14. NO. OFPRINTED PAGES
72
15. Price
USCOMM-DC e043-PS0
U.S. GOVERNMENT PRINTING OFFICE 1988—574-791 85.109 REGION NO 8
NBS_ Technical Publications
Periodical
Journal of Research—The Journal of Research of the National Bureau of Standards reports NBS research
and development in those disciplines of the physical and engineering sciences in which the Bureau is active.
These include physics, chemistry, engineering, mathematics, and computer sciences. Papers cover a broad
range of subjects, with major emphasis on measurement methodology and the basic technology underlying
standardization. Also included from time to time are survey articles on topics closely related to the Bureau's
technical and scientific programs. Issued six times a year.
Nonperiodicals
Monographs—Major contributions to the technical literature on various subjects related to the Bureau's scien-
tific and technical activities.
Handbooks—Recommended codes of engineering and industrial practice (including safety codes) developed in
cooperation with interested industries, professional organizations, and regulatory bodies.
Special Publications—Include proceedings of conferences sponsored by NBS, NBS annual reports, and other
special publications appropriate to this grouping such as wall charts, pocket cards, and bibliographies.
Applied Mathematics Series—Mathematical tables, manuals, and studies of special interest to physicists,
engineers, chemists, biologists, mathematicians, computer programmers, and others engaged in scientific andtechnical work.
National Standard Reference Data Series—Provides quantitative data on the physical and chemical properties
of materials, compiled from the world's literature and critically evaluated. Developed under a worldwide pro-
gram coordinated by NBS under the authority of the National Standard Data Act (Public Law 90-3%).
NOTE: The Journal of Physical and Chemical Reference Data (JPCRD) is published quarterly for NBS bythe American Chemical Society (ACS) and the American Institute of Physics (AlP). Subscriptions, reprints,
and supplements are available from ACS, 1155 Sixteenth St., NW, Washington, E)C 20056.
Building Science Series—Disseminates technical information developed at the Bureau on building materials,
components, systems, and whole structures. The series presents research results, test methods, and perfor-
mance criteria related to the structural and environmental functions and the durability and safety
characteristics of building elements and systems.
Technical Notes—Studies or reports which are complete in themselves but restrictive in their treatment of a
subject. Analogous to monographs but not so comprehensive in scope or definitive in treatment of the subject
area. Often serve as a vehicle for final reports of work performed at NBS under the sponsorship of other
government agencies.
Voluntary Product Standards—Developed under procedures published by the Department of Commerce in
Part 10, Title 15, of the Code of Federal Regulations. The standards establish nationally recognized re-
quirements for products, and provide all concerned interests with a basis for common understanding of the
characteristics of the products. NBS administers this program as a supplement to the activities of the private
sector standardizing organizations.
Consumer Information Series—Practical information, based on NBS research and experience, covering areas
of interest to the consumer. Easily understandable language and illustrations provide useful backgroundknowledge for shopping in today's technological marketplace.
Order the above NBS publications from: Superintendent of Documents, Government Printing Office,
iVashington, DC 20402.
Order the following NBS publications—FIPS and NBSIR 's—from the National Technical Information Ser-
vice, Springfield, VA 22161.
Federal Information Processing Standards Publications (FIPS PUB)—Publications in this series collectively
coastitute the Federal Information Processing Standards Register. The Register serves as the official source of
information in the Federal Government regarding standards issued by NBS pursuant to the Federal Property
and Administrative Services Act of 1949 as amended, Public Law 89-306 (79 Stal. 1 127), and as implcmcniedby Executive Order 1 1717 (38 FR 12315, dated May 11. 1973) and Part 6 of Title 15 CFR (Code of Fcxleral
Regulations).
NBS interagency Reports (NBSIR)—A special series of interim or final reports on work pcrfomicd by NBSfor outside sponsors (both government and non-government). In general, initial distribution is handled by the
sponsor; public distribution is by the National Technical Information Service, Springfield, VA 22161, in paper
copy or microfiche form.
U.S. Department of CommerceNational Bureau of Standards
Gaithersburg, MD 20899
Official BusinessPenalty for Private Use $300